Last night I read Scott Soames' article The Gap Between Meaning and Assertion. According to Soames, the assertoric content of an utterance of a sentence is (roughly) the proposition the speaker intended to assert; and the semantic content of a sentence in context is (roughly) the semantic values of the constituents of the sentence relative to the context in question. Soames thinks that the speaker might assert the semantic content as well as the assertoric content. This is so if the semantic content is a "relevant, unmistakable, necessary and a priori consequence of the speaker's primary assertions, together with salient presuppositions of the conversational background". Let us consider an example (from Soames).
Peter doesn't realize that Paderewski, the statesman, is Paderewski, the musician.
(1) I believe that Paderewski has great musical talent (said by Peter to fellow musicians)
(2) I don't believe that Paderewski has great musical talent (said by Peter to fellow political friends).
According to Soames, here is what Peter asserted in the two cases:
(1a) I believe that Paderewski, the musician, has great musical talent.
(2a) I don't believe that Paderewski, the statesman, has great musical talent.
However, Soames says, when Peter utters (1) he is also asserting the semantic content of (1). For "it is an obvious necessary and a priori consequence of the fact that one believes of a man that he has the property of being both A and B, that one believes, of him, that he has the property of being B"
I am not sure I follow. 'S believes that x is A and B' doesn't entail 'S believes that x is B', for 'believe' is not closed under logical consequence. Is the latter a "relevant, unmistakable, necessary and a priori consequence of the former"? Well, doesn't that depend on who the speaker is, and how complicated the predicates 'A' and 'B' are?
Tuesday, August 01, 2006
Semantic Content and Assertoric Content
Subscribe to:
Post Comments (Atom)
7 comments:
I'm not sure I see why there's such a heavy play with believe contexts in the examples (1) and (2). Does Soames point fall apart if the utterances are (contexts the same as in (1) and (2)):
(1*) Paderewski has great musical talent
(2*) Paderewski has not got great musical talent,
with suitable alterations made to (1a) and (2a)? Then the semantic content (1*) will be an easy consequence of the assertoric content (1a*), skirting the issue over whether belief is closed under such a consequence relation.
(I recognise that this is bound to be one of those 'should have read the paper in question before commenting' comments. But hey, it's late.......)
Belief is not closed under known logical consequence, but all that says is that for some P and Q where P entails Q, S can believe P and know that P entails Q without believing Q. It doesn't say that closure fails with respect to every two propositions where one entails the other. I find the entailment from 'S believes that A & B' to 'S believes that A' enormously plausible. It doesn't matter whether S is irrational or whether A and B are very complicated. If S is rational enough and attentive enough to believe A & B, then he thereby believes A. He doesn't have to draw some inference to get to A, opening the door to closure failure; simply by believing A & B, he believes A.
If that's wrong, is there a plausible case where S believes A & B but doesn't believe A?
Sorry, I was being sloppy by not distinguishing the entailment from S believes that A & B to S believes that A and the entailment from S believes that x is A & B to S believes that x is A, which is what you are talking about. Still, I stand by what I said above with respect to this entailment. It seems that by believing that x is A & B, S thereby believes that x is A. I'd be interested to see a plausible case where S believes x is A and B but doesn't believe x is A.
Aidan: Soames' view doesn't apply just to belief reports. I just picked that example. Whenever the semantic content is an obvious necessary and a priori consequence of the assertoric content, then the semantic content counts as asserted as well.
Leo: you're right that it is very plausible that 'S believes x is A and B' entails 'S believes that x is B'. I guess I was thinking that it may be that S would assent to 'x is A and B' and yet would not assent to 'x is B'. But without the time leap, the inference is very plausible.
Interestingly, 'S desires that p and q' does not entail 'S desires that p'. For example, I might desire that John and Mary attend the party, because together they are so much fun, and yet fail to have any desire (one way or the other) with respect to Mary's attendance. Perhaps Mary is no fun at all when she is on her own. So if you were to ask me 'do you want Mary to attend the party?' I might reply 'that depends'.
Yes, I got that, I guess I was thinking that if we're just looking for an example, the non-belief cases looked more promising.
Kelp and Pritchard offer counterexamples to the distribution of belief over conjunction (in the paper Salerno links to in his recent post on Church's solution to the knowability paradox) which have basically the same shape as the example you give in the belief case. I'm firmly with orthodoxy on this matter, but if they're right belief and desire are pretty much alike wrt distribution.
"which have basically the same shape as the example you give in the belief case".
Urg, I meant desire here.
Hi Aidan
Yes, thanks for reminding me! I read the Duncan/Kelp paper a while ago, but it is obviously relevant here.
Post a Comment